Installation of the Python interface

Note: The following steps are necessary if you intend to use the CommonRoad Drivability Checker from Python. They are not necessary if you only intend to use the CommonRoad Drivability Checker from C++.

In order to install CommonRoad Drivability Checker, we first need to install essential third party libraries, followed by installation of optional third party libraries, and finally the installation of CommonRoad Drivability Checker itself. All required steps are described in the following sections.

We provide two installation methods:

  • Method 1: Automatic Installation via Bash Script (recommended for Ubuntu)

  • Method 2: Manual Installation

Before installing anything, we recommend you to install Anaconda and to create an Anaconda environment. Below, we assume that your Anaconda environment is called commonroad-py36. In the root folder of the CommonRoad Drivability Checker, activate your environment with:

$ conda activate commonroad-py36

If you choose to continue with python virtualenv please refer to the official documentation, and create an environment with python3.6m (dev version), and activate you environment.

Method 1: Automatic Installation via Bash Script

Note: The installation script only works on Ubuntu based distributions. If your operating system is not Ubuntu based, please refer to Method 2: Manual Installation, which is described below.

We have provided a build script for easy installation. You still need to activate your conda environment (or virtual environment if you like). Full installation option will install third party libraries (suggested).

The -i option on the build command will install the drivability checker to the environment specified in the path.

  1. Open your console in the root folder of the CommonRoad Drivability Checker.

  2. Activate your conda environment with the following command (if not already activated)

    $ conda activate commonroad-py36

  3. Run the build script

    Basic installation without CGAL

    $ bash build.sh -j JOB_COUNT -i

    (SUGGESTED) Full installation:

    $ bash build.sh -j JOB_COUNT --cgal -i
    Note that you have to replace

    • JOB_COUNT with the number of jobs you are willing to allocate to cmake, for example -j 2. Each job (possibly) will use a core, so specify this number according to your system and free cores.

    For additional options, please run bash build.sh -h command to view them.

Method 2: Manual Installation

Prerequisites

We assume that the following libraries are already installed on your system:

  • Eigen3 (version >=3.3.7 - preferably the newest version)

  • Boost libraries (version >= 1.58)

  • OpenMP (for the geometry submodule support)

For the documentation, we require the libraries Pandoc and Doxygen. All aforementioned libraries can be installed on Ubunutu via apt-get.

Installation of Essential Third Party Libraries and Packages

The following dependencies are not included in this repository and need to be installed externally, e.g. by using a package manager:

  • Boost (only Boost.Thread, Boost.Test, Boost.Filesystem)

  • Eigen

Installation on Linux (e.g. using apt-get for Debian derivatives):

$ sudo apt-get install build-essential cmake git wget unzip libboost-dev libboost-thread-dev libboost-test-dev libboost-filesystem-dev libeigen3-dev libomp-dev

Installation on macOS using Homebrew:

$ brew install cmake eigen boost

On macOS, it is also necessary to install the OpenMP library manually. The version of the OpenMP library must correspond to the version of the Apple C++ compiler currently installed on your Mac (g++ –version). One can download the corresponding version of the library from https://mac.r-project.org/openmp/ and follow the installation instructions.

The following third party libraries are included as submodules:

In order to initialize the bundled submodules, run the following commands in the root folder of the CommonRoad Drivability Checker:

$ git submodule update --init

Following packages are available via PyPi:

They can be installed with the following command:

$ pip3 install -r requirements.txt

N.B. If you are using Conda, make sure your environment is activated!

Installation of Optional Third Party Libraries

(Recommended) Some optional functions for polygon triangulation require a third-party library Triangle to be installed. To do so, run

$ pip3 install triangle

Please note that although Triangle is freely available, it is copyrighted by its author and may not be sold or included in commercial products without a license. See the original licensing conditions of the Triangle library as well as the Triangle website for details.

It is possible to modify Drivability Checker to support CGAL. For the installation of CGAL, please refer to their website.

Installation of the CommonRoad Drivability Checker

After installing all essential third party libraries and packages, you can now install the CommonRoad Drivability Checker.

  1. Open your console in the root folder of the CommonRoad Drivability Checker.

  2. Activate your conda environment with

    $ conda activate commonroad-py36

  3. Please make sure that the latest pip version is installed. Otherwise, the next installation step would not work.

    $ python -m pip install --upgrade pip

  4. Compile and Install the CommonRoad Drivability Checker library by running

    $ BUILD_JOBS=8 python setup.py build
$ pip install .

    Note: This will automatically build all C++ dependencies of the CommonRoad Drivability Checker. The number 8 in this example indicates the number of CPU cores to be used for the compilation. Each job (possibly) will use a core, so specify this number according to your system and free cores. Advanced build settings can be modified in the file setup_options.py.

Canceling the build and then restarting it should generally be safe, however make sure that the Python environment you activated stays the same. In case of any errors, try deleting the `build folder and running pip again.